Unsupervised traffic abnormal behavior detection method based on foreground target detection

The invention discloses an unsupervised traffic abnormal behavior detection method based on foreground target detection, and the method comprises the following steps: rapid target detection, appearance abnormal feature extraction, track abnormal feature extraction, abnormal feature detection and processing of other conditions. The detection of a target vehicle on an expressway is carried out through a YOLO v5 model. The complex calculation amount of directly extracting features from information such as low-layer optical flow and gradient is avoided, the robustness and accuracy of feature extraction from high-dimensional semantic information are ensured, local feature information of a target in a video is extracted, hidden features in depth information are extracted through an auto-encoder to perform anomaly detection, depth trajectory information is directly extracted from a current frame and front and back continuous frames through 3D convolution, and hidden features in the depth information are extracted th